Spatial and temporal characteristics and driving mechanisms of recreational ecosystem services supply-demand mismatch in rapidly urbanizing areas: Evidence from the Yangtze River Delta

• Updated framework for evaluating the demand and supply of recreational ecosystem services at the urban agglomeration scale. • Changes in the matching of supply and demand are characterized by an urban–rural stratification. • Deepening of supply–demand conflicts and widening of conflict regions in 2010–2020. • Quantitatively analyzed the spatial spillover and path-locking effects of shifting supply and demand states. • Effectively identifying non-stationary, non-linear and thresholds in the spatial manifestation of natural and human drivers. The ecological changes triggered by rapid urbanization pose a significant threat to the balance of supply and demand for recreational ecosystem services. Understanding the spatial and temporal processes of Recreational Ecosystem Services (RES) supply–demand matching and their driving mechanisms is essential for the enhancement of human well-being and the sustainable management of socio-ecological systems in the region. Taking the Yangtze River Delta Urban Agglomeration in China as an example, this study first builds an updated supply indicator system based on the four dimensions and examines the supply, demand, and matching of RES in the region with the help of heterogeneous data from multiple sources. We also examine the spatio-temporal processes and spatial spillover effects of supply–demand matching, as well as the nonlinearity and spatial non-stationarity of the role of the driving factors. The results show that the supply level of RES is generally low, but is on an upward trend. There are significant spatial differences in the demand level, and the matching level is declining with significant polarisation. The matching types generally show a trend of decreasing L-L and M-L and a significantly increasing H-H. The changes in supply and demand risks at different levels are dominated by level-by-level migration, but the low-level units attract the higher level to leap downward across level transitions, thus indicating a significant spatial spillover effect. The regression model estimation showed a decrease in the contribution of ecological factors and an enhancement of socio-economic factors. Each factor exhibits different degrees of non-linear marginal effects. The regression coefficients were significantly heterogeneous in space. Finally, the influence processes of county-level and city-level drivers were discussed, and the establishment of a three-level transmission mechanism of “grid-county-city” was proposed to achieve the sustainable management of recreational resources.